Lies, damned lies, and GAA stats

There has been much discourse of late about the lack of crowds at the football league finals. No one, that I have seen, however has any empirical data to show whether it is a mere year on year blip, caused by Dublin being in the finals in 2011, or an actual downward trend.

The reason I start my first post on this subject is because it is recent evidence of opinion becoming overnight fact. The GAA is awash with quick, lazy analysis that is not backed up by any data. Compare the lack of availability of very basic statistics – possession %, turnovers, shot rates, conversion rates etc. – against the near certain pronouncements that certain teams or players are good/bad.

This is nothing new. Back in 2010 I got so annoyed at the lack of statistical data that I followed in the footsteps of one of my favourite websites and created my own database. A little aside

I watch a lot of NFL and one of the best sites is www.footballoutsiders.com. They originally set out to debunk the widely held view that you have to run the ball to win … they then progressed to produce some of their own stats as whilst the NFL box scores are useful they can lead to meaningless stats – a 5 yard run on 3rd & 4 is worth a hell of a lot more than a 5 yard run on 4th & 10. The box scores, and cumulative totals, hide this. So what they did was break down every possible pass & run on every down, get an average and then compare every team’s return on every run/pass to this average. Now you had a better ranking system

Oh what I’d give for box scores on GAA!! Try finding something as simple as the top point scorer from play in the Championship. It’s nowhere. In creating the database that’s what I focussed on – shooting & scoring. Who shot from where, with what frequency and with what accuracy?

The reason for focusing on shooting and offence, rather than defence, is that the outcome is attributable to one player and one area of the pitch. Shooting is a skill – its accuracy is fundamental to the game and more importantly measurable. In today’s game it is very hard to attribute one forward to one back. If Hughie McGrillen attacks for Kildare, it breaks down and the corner forward scores on the counter attack, does the ‘blame’ lie with McGrillen or his team-mates for not covering. If so which team-mate? Now we can infer some defensive statistics from plotting scores against a team however the focus is on attacking plays.

I charted every possession and every shot from 36 games on TV – I made up my own database, got the averages, and ranked players and teams accordingly. Now there are obvious issues in that

(a) I have half the season and

(b)  there would be bias towards the teams that were on TV more often

but in the end I had over 2,000 shots and 3,000 possessions from 36 games. Any bias was diluted enough by volume to the extent that the findings were plausible/reliable.

Due to time and general ‘having two small kids’ issues I didn’t follow up in 2011 however I intend to track 2012’s Championship and hopefully we’ll be able to dispel some of the talking heads’ bluff and bluster. In the run up to the Championship I’ll post some of my findings from 2010.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: